Taylor Swift Fights Back: New Trademarks Target AI Deepfakes

Taylor Swift has moved to trademark elements of her voice and image in a bid to strengthen legal protections against artificial intelligence deepfakes, according to filings with the U.S. Patent and Trademark Office.

The applications were filed on Friday, April 25, 2026, by TAS Rights Management. They cover two audio clips of Swift’s voice and one stage image, aiming to limit unauthorized uses that mimic her identity.

U.S. trademark filings target voice and stage image

The filings are framed as a defensive step against AI-generated content that can replicate a performer’s voice or likeness. The goal, as described in the source material, is to create additional legal safeguards that could be used when content is manipulated or generated to appear as if it came from Swift.

The approach reflects a concern that existing legal tools may not fully address how modern AI systems can produce new material that sounds like an artist without directly copying a specific recording.

Why trademarks are being used alongside copyright

Copyright law has traditionally been used to protect recorded music. However, the source material notes that AI technologies can generate new content that imitates an artist’s voice without reproducing the underlying sound recording itself.

By seeking trademark protection for specific audio clips and a distinctive visual image, the filings could create a different legal basis for challenging unauthorized uses. In practice, this would be separate from arguments focused on copying of existing recordings.

Legal experts flag a new and untested strategy

Trademark attorney Josh Gerben said the applications are specifically designed to protect Swift from AI-related threats. He described the filings as a targeted response to the risks posed by deepfakes and other forms of synthetic media.

The source material also points to “Right of Publicity” laws as an existing line of defense against unauthorized use of a famous person’s likeness. The trademark strategy is presented as an additional layer that could sit alongside those protections.

At the same time, the effort is described as novel. The registration of a celebrity’s spoken voice as a trademark has not yet been tested in courts, leaving uncertainty about how judges would treat such claims if challenged.

The filings underscore how artists and rights holders are exploring new legal pathways as AI tools make it easier to generate convincing imitations. For now, the outcome will depend on how the trademark applications proceed and how enforceable such protections prove to be in real disputes.